Data Envelopment Analysis (DEA)

for Services Productivity

Lecture 2, Prof Pradeep Pai

1
Solver Example
Product Product Product Available
Machines A B C

I 4 5 3 90

II 3 2 3 54

III 2 4 3 124

Profits 40 60 38

2
 Data Envelopment Analysis (DEA).
 How can corporate management evaluate productivity of a bank
branch?
 Difficulties are three fold,
 Appropriate inputs & measures of these inputs.
 What are the appropriate outputs.
 Appropriate ways of measuring the relationships between inputs &
outputs.
 The problem with simple ratios is that the mix of outputs or inputs
is not considered explicitly.
 Broad based measures such as profitability or return on
investments are highly relevant as overall performance measures,
but inefficient to evaluate operating efficiency of a single service
unit. A profitable bank branch need not be necessarily efficient!!
 DEA measure of efficiency explicitly accounts for a mix of outputs
and inputs and is therefore more comprehensive and reliable than a
set of operating ratios or profit measures.
 We will use the Solver Sensitivity Analysis and the concept of
Shadow price
3
DEA Formulation
 Objective Function: Maximize Output,
Max Z = W1u1 + W2u2 + W3u3 + W4u4 +
….+ Wnun
 Constraints: Common constraints (as
many DMU’s) + Unique constraint.
 Common Constraints: Outputs –
Inputs ≤ 0 orui – vi ≤ 0
 Unique Constraint: For DMU being
considered, v1 + v2 + v3 + v4 + …. = 1
4
Burger Palace.
Service Meals Labour Material
Unit Sold Hours Dollars
(u1) (v1) (v2)
1 100 2 200
2 100 4 150
3 100 4 100
4 100 6 100
5 100 8 80
6 100 10 50
5
Burger Palace - Notations.
Service Meals Labour Material
Unit Sold Hours Dollars
(u1) (v1) (v2)
S1 100 2 200
S2 100 4 150
S3 100 4 100
S4 100 6 100
S5 100 8 80
S 100 10 50 6
 Summary of DEA results.
Service Efficiency Relative Relative
Unit Rating LabourHour Material
Value(v1) Value(v2)
S1 1.000 0.1667 0.0033
S2 0.857 0.1428 0.0028
S3 1.000 0.0625 0.0075
S4 0.889 0.0555 0.0067
S5 0.901 0.0568 0.0068
S6 1.000 0.0625 0.0075
7
Graph of Productivity Frontier
Productivity Burger Palace

250

200 2, 200
Material Dollars

150 4, 150

100 4, 100 6, 100

8, 80

50 10, 50

0
0 2 4 6 8 10 12
Labour Hours

8
Analysis for S4
Productivity Burger Palace

250

200 2, 200
Material Dollars

150 4, 150

100 4, 100 6, 100

8, 80

50 10, 50

0
0 2 4 6 8 10 12
Labour Hours

9
Analysis for S2
Productivity Burger Palace

250

200 2, 200
Material Dollars

150 4, 150

100 4, 100 6, 100

8, 80

50 10, 50

0
0 2 4 6 8 10 12
Labour Hours

10
Analysis for S5
Productivity Burger Palace

250

200 2, 200
Material Dollars

150 4, 150

100 4, 100 6, 100

8, 80

50 10, 50

0
0 2 4 6 8 10 12
Labour Hours

11
DEA Formulation
 Objective Function: Maximize Output,
Max Z = W1u1 + W2u2 + W3u3 + W4u4 +
….+ Wnun
 Constraints: Common constraints (as
many DMU’s) + Unique constraint.
 Common Constraints: Outputs –
Inputs ≤ 0 orui – vi ≤ 0
 Unique Constraint: For DMU being
considered, v1 + v2 + v3 + v4 + …. = 1
12
 NMIMS productivity
improvement (not actual)
Division Student with A Grade Good Placements Teaching Hours Practical Hours

A 30 25 40 35
B 28 26 38 33
C 27 29 36 37
D 31 20 37 32
E 30 27 35 35
F 32 22 38 33
G 29 28 31 34
H 27 30 32 30
I 26 31 38 32
J 31 28 36 33
13
DEA Strategy Matrix
Efficiency
Low High

Understan
d why they Benchmar
High

are k Group
profitable
Profit

Candidate
Problem
Low

s for
Branches
Divestiture

14
Benefits of DEA
 Circumvents the need to have standard costs.
 Incorporates multiple inputs & outputs in the
analysis without a need for conversion into a
common dollar basis.
 Inefficient units can be identified, magnitude
of inefficiencies can be measured, and
discover ways to reduce the inefficiencies.
 Comparison is with groups of similar service
units delivering the same services.
 Uses the concept of RELATIVELY efficient
unit.
 Provides targets for inefficient units to
become efficient 15
 Mid-Atlantic Bus Lines.
Bus City Served Ticket Freight Labour Facility
Depot Sales Sales Hours Dollars
1 Philadelphia 700 300 40 500
2 Baltimore 300 600 50 500
3 Washington 200 700 50 400
4 Richmond 400 600 50 500
5 Raleigh 500 400 40 400
6 Charleston 500 500 50 500
7 Savannah 800 500 40 600
8 Jacksonville 300 200 30 400
16
 Mid-Atlantic Bus Lines.
Bus City Served Eff Relative Relative
Depot Labour Facility
1 Philadelphia 1 0 0.002
2 Baltimore 0.8737 0.02 0
3 Washington 1 0 0.0025
4 Richmond 0.9022 0.00255 0.00175
5 Raleigh 1 0.0032 0.0022
6 Charleston 0.8936 0.0025 0.0017
7 Savannah 1 0 0.00167
8 Jacksonville 0.5714 0.0033 0.0023
17
Analysis for Depot 8,
Jacksonville
 Bus Depot 8 is between Bus Depot 5 & Bus
Depot 7.
 0.1429x40 + 0.2857x40 = 17.144, which
means bus depot 8 is using excess labour
hours 30 – 17.144 = 12.856 hours.
 0.1429x400 + 0.2857x600 = 228.58, which
means bus depot 8 is using excess facility
dollars by 400 – 228.58 = $171.42.

18